Skip to main navigation Skip to search Skip to main content

Noise-assisted Nonlinear Squeezing Transform for Rotating Machinery Fault Diagnosis

  • Xi'an Jiaotong University
  • Case Western Reserve University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper proposes a noise-assisted nonlinear squeezing transform (NANST) technique based on the spectrogram zeros (SPZ). This method can effectively weaken the relative amplitude of noise and thus enhance signal characteristics for rotating machinery fault diagnosis. The statistical characteristics of the zeros are analyzed, and it is found that the zeros of noise have random distribution characteristics, and the signal of interest has fixed structure and fixed zeros. Thus, the results of multiple nonlinear squeezing transform (NST) results with different noises are fused by the geometric mean to average and weaken the noise effect. Therefore, the results with clearer time-frequency representation (TFR) can be obtained, even for weak signals. The performance of the proposed method is verified in the rotating machinery high-speed bearing experiment.

Original languageEnglish
Title of host publicationIEEE International Instrumentation and Measurement Technology Conference, I2MTC 2025 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798331505004
DOIs
StatePublished - 2025
Event2025 IEEE International Instrumentation and Measurement Technology Conference, I2MTC 2025 - Chemnitz, Germany
Duration: 19 May 202522 May 2025

Publication series

NameConference Record - IEEE Instrumentation and Measurement Technology Conference
ISSN (Print)1091-5281

Conference

Conference2025 IEEE International Instrumentation and Measurement Technology Conference, I2MTC 2025
Country/TerritoryGermany
CityChemnitz
Period19/05/2522/05/25

Keywords

  • Fault diagnosis
  • Time-frequency analysis
  • Zeros of the spectrogram

Fingerprint

Dive into the research topics of 'Noise-assisted Nonlinear Squeezing Transform for Rotating Machinery Fault Diagnosis'. Together they form a unique fingerprint.

Cite this